IBM 5110-3 / 5120 8inch drives, disk timing error and Process Check issue

[voidstar78]

So - I got this 5120 booting up to BASIC. It runs program fine, I think all the RWS/RAM is fine.

But I still can't access disks. I got a set of DS/DD to try - the index hole on them seems to be in the same location as the one on the original IBM 8" disk that came with the system.

So I've uncaged the drives and took a look inside. The circuit boards don't look all that dusty. Cables look ok, central cylinder seems to be spinning fine.







I get different error numbers. Most often I am getting
ERR 001 Diskette timing error
then also... sometimes I get one of these:
ERR 013 Device not attached
ERR 014 Device not selected

If I install the "terminator" to the back of the system, then I always get ERR 045 during startup: Diskette inserted incorrectly, or no diskette present. (so I have to insert a disk and close the door during startup, only if the "terminator" is attached at the IO block at the back of the system).

NOTE: I noticed the 8" disks "wobble" when inserted and the system is powered on. Working on a short video of that to show what I mean. It's not super excessive, but noticeable. Should 8" disks that are inserted (but spinning) be super steady? or is it normal for them to "wobble" a little bit?


And one other thing: even after "booting" to BASIC and running simple prorgrams, the red "Process Check" light is always on as soon as I power up the system. The system still proceeds with the startup sequence (ROS CRC checks, etc.)

Logically, whatever problem that is causing that should be fixed before investigating more about the disk drive issues. I tried going through the MAP sequences, which has a "Is Process Check light on?" sequence: MAP 420 is all about the Process Check! I eventually get down to this part, where it ask "Is the UP light on?" Which is referring to the CE logic probe described in MIM appendix B. I'm just getting a trickle of voltage (~.15V) on that pin, which I'm not sure how to interpret (i.e. IBMs general logic probe would interpret that as "UP" or not?). But to note, all the keys on the keyboard seem fine - so I don't think its KBD parity check.




I've not messed with 8" drives before. In the first image above, the little lever near the bottom left is always propped up -- sometimes the system will actuate it down, but then it always pops back up fairly immediately. // And I haven't had to diagnos Process Check light issues before.


Open to suggestions before giving up on these disk drives. The same cards I'm using on this 5120/5110-3 I can also swap over to a 5110 - and that 5110 will startup fine with no Process Check light staying on. Well, or at least I'll double check on that tomorrow -- that is, I need a way to induce a Machine Check issue on that original 5110 to double check and make sure its red LED light is still working (or I guess I can just apply my own power on it for a bit).


-v*

 

[voidstar78]

Sample of bench testing the disk drive. The head "clicks"/clamps(?) down during startup, but then retracts back right away. Maybe that's normal?

And not sure if the "vibration" of the disk (shown towards end of clip) is normal.

 

[tradde]

To me the head seems to not be on the disk long enough to do much. Does it even have time to read the sector number? I am not a floppy expert and never worked with 8 inch at home. What happened to the front doors? I think they might prevent that wobble so that might be Ok as you have it. Is it causing an issue? Maybe. I'd clean up the slide rails and lube them for sure. Seems to be a bit of rust/corrosion on some parts. I would also expect the head to move forward or backward a bit too. Which it does not seem to do. Where is track zero on an 8" drive? On the outer edge where it is? Or in towards the center?

 

[voidstar78]

All good questions. And same, my first time working with 8" drives as well.

Looks like the disk drive hardware on the 5120 is a bit different than the 5114 external drives. 5120 disk drive section is 2-19 of this at bitsavers:

http://www.bitsavers.org/pdf/ibm/5120/SY34-0192-0_5120maint_Dec79.pdf

But it hasn't been processed to be searchable, so a little slow going.

One thing I came across, there are some additional fuses "behind" the power supply. The one fuse I found behind the CRT was blown on this 5120 - so I better go check those other fuses as well. Physically (manually when powered off), yes the head will move forward and back. The disk is spinning so clearly getting some power, but still worth checking the fuses. It states that I have to remove the keyboard to do so (section 2-80), but I think I can also just slide the entire power supply out - but won't get it to that till the weekend.

As to the front doors: there aren't "full cover" doors like there was in the 5114 external drives used on the IBM 5110. There are "latches" on both bays, and those seem to be working fine -- the drive is large and those latches are just at the very edge of the video. Per section 3-2, looks like the older 5114 drive units can also work on the 5120 also.

Page 3-23 should answer the track zero question:


Lots to read up on... But will start by checking the disk drive fuses (I thought the system just had that one main fuse behind the CRT).

 

[stepleton]

The YouTube channel of Jerry Walker / JM Precision has a few videos of a repair attempt on a seriously damaged 5120. One or two of the episodes are dedicated to the floppy drive, but more to fixing them mechanically than anything. Still, you may find some useful info...

 

[voidstar78]

Uh oh, apparently one has to "low level format" these disks first using a program called INITIAL (and declare a desired format, indexed "1-9"). Only after that can they be MARK'd for use on the 5110/5120 systems. So, I do have one original IBM disks, although I don't know if it's still "working" (i.e. no bad sectors, content intact, etc), but not sure if it has that INITIAL program.

Will probably put the "disk project" on hold, as a few other things to focus on first.

 

[voidstar78]

So, the manual isn't kidding, getting to the internal fuses is easier to just remove the keyboard. But fortunately, that's easy, the keyboard is just two screws.

The internal fuses all look fine to me.





Also I noticed at Norbert's 5110 emulator online, the virtual disk mounted seems to include the INITIAL program.



And sure enough, Corti's original 5110 emulation includes a csf.dsk and csf.tap (customer support functions). So, there goes my excuse to give up But how to transfer the disk image to a physical disk, hmm. Or if I can get the binary of what INITIAL is, I could just stream it into memory manually and run it, perhaps.